This dissertation consists of three essays concerning inflation forecasting, taking the Brazilian case as an application. In the first essay, we examine the effectiveness of several forecasting methods for predicting inflation, focusing on aggregating disaggregated forecasts. We consider different disaggregation levels for inflation and employ a range of traditional time series techniques, as well as linear and nonlinear machine learning (ML) models that deal with a larger number of predictors. For many forecast horizons, aggregation of disaggregated forecasts performs just as well as survey-based expectations and models generating forecasts directly from the aggregate. Overall, ML methods outperform traditional time series models in predictive accuracy, with outstanding performance in forecasting disaggregates. In our second essay, we investigate the potential benefits of combining individual inflation forecasts by proposing a time-varying bias correction for the average forecast. Our analysis includes estimations using both rolling windows and state-space models that use the recursiveness of the Kalman filter. We achieve good forecast performance for models based on small rolling windows for shorter and intermediate forecast horizons, while a state-space model performs slightly worse than procedures based on rolling windows. In the third essay, we use supervised learning to generate forward-looking indexes based on tweets and news articles for accumulated inflation and investigate whether these indexes can improve inflation forecasting performance. Our results indicate that news-based indexes provide significant predictive gains, particularly for 3- and 12-month-ahead horizons. These findings suggest that incorporating more information sources than just expectations based on experts’ opinions can lead to more accurate forecasts.

This paper investigates the impacts of neighborhoods on the economic outcomes of adults. We exploit one of the world's largest housing lottery programs and administrative data linking lottery registration, formal employment, and access to social programs in Brazil. Receiving a house has positive impacts on housing quality and reduces household expenditures but has negative effects on beneficiaries' neighborhood characteristics. On average, the program has a  negative impact on the probability of being formally employed but no effects on the quality of jobs. Poorer individuals, however, experience better formal employment outcomes and lower welfare dependency. We find no differential impacts by distance to beneficiaries' previous homes or jobs. Leveraging a double-randomization design to allocate houses, we show that there are significant differences in effects across neighborhoods and we propose a framework to estimate the relative importance of potential underlying mechanisms. Network quality, amenities and crime play a very limited role, while labor market access explains 82-93% of the observed differences in neighborhood effects.

European macroeconomic and financial aggregates move in lockstep over the business cycle. We develop a model in which a single risk premium shock triggers these comovements. The key feature is a financial sector where traditional banks transfer part of their risky loan portfolio to nonbank institutions. We fit the model to euro area data and find that risk premium shocks are the main driver of business and financial cycles over the past two decades.

We study a unique dataset with comprehensive coverage of daily prices in large multiproduct retailers in Israel. Retail stores synchronize price changes around occasional “peak” days when they reprice around 10% of their products. To assess aggregate implications of partial price synchronization, we develop a new model in which multiproduct firms face economies of scope in price adjustment, and synchronization is endogenous. Synchronization of price changes attenuates the average price response to monetary shocks, but only high degrees of synchronization can substantially strengthen the real effects of monetary policy shocks. Our calibrated model generates real effects similar in magnitude to those in Golosov and Lucas (2007).

This paper investigates the impact of public provision on competition within the Brazilian higher education sector. We develop and estimate an empirical model of demand for higher education that incorporates tuition-free institutions and consumer choice constraints. Our model produces more realistic substitution patterns than a logit model without constraints, indicating that selectivity is an essential dimension of product differentiation. We find that the most selective public programs exert comparable competitive pressure to the most selective private programs, but the least selective public programs exert more competitive pressure than the least selective private programs. Our estimates of the supply response of private institutions suggest that, in the absence of public programs, tuitions would be about 7 percent higher. These findings provide important insights into the competitive dynamics of Brazilian higher education and highlight the role of public provision in promoting competition in this sector.

Does fiscal policy risk affect the yield curve in an emerging economy? How can we adequately measure this kind of uncertainty? Exploiting the case of Brazil, we estimate a novel, news-based measure of fiscal policy risk using natural language processing. We show that increases in fiscal policy risk are associated to increases in the levels of long maturities in the yield curve, in the term spread and to a depreciation of the exchange rate. The effects are robust to a series of alternative specifications of the text-based index, suggesting that fiscal risk is a relevant phenomenon in the Brazilian setting.

It is a consensus in economic theory that the increase in the terms of trade leads to an appreciation of the real exchange rate. However, during the recent period of the COVID-19 pandemic, this relationship appears to have been disrupted, as there has been a significant rise in commodity prices but the real exchange rates of many countries have not appreciated correspondingly. The aim of this M.Sc. Thesis is to examine the reasons for this deviation from the established correlation. I estimate several SVARs for commodity exporting countries with a recursive block identification scheme and conclude that structural shocks other than the commodity one explained the real exchange rate depreciation in the pandemic period. In 2020, the global risk was the main factor responsible for depreciating the exchange rate, while in 2021 the high country risk, specially for emerging countries, and the low level of the domestic interest rate appear as the main factors responsible for this break.

We study the implications of transparency policies on decentralised public good provision by the non-profit sector. We present a model where imperfect monitoring of the use of funds interacts with the competitive structure of the non-profit sector under alternative informational regimes. Increasing transparency regarding the use of funds may have ambiguous effects on total public good provision and on donors’ welfare. On the one hand, transparency encourages all non-profit firms to engage more actively in curbing fund diversion. On the other hand, it tilts the playing field against non-profits facing higher monitoring costs, pressing them to give up on their missions. This effect on the extensive margin implies that transparency policies lead to a reduction in the diversity of social missions addressed by the non-profit sector. We show that the negative impact of transparency on social mission variety and on donors’ welfare is highest for intermediate levels of asymmetry in monitoring costs.

This dissertation investigates the impact of environmental sanctions on forest changes in the Brazilian Amazon. We examine the role of general deterrence in changing farmers’ behavior by altering their perceived risk of violating forest laws after being punished or witnessing the punishment of adjacent farms. Using a difference-in-differences strategy and novel spatial data, we show that sanctions decrease deforestation by 48% in punished farms and 21% in adjacent farms while increasing reforestation by 15% and 6.5%, respectively. Heterogeneity analyses reveal that even sanctions lacking incapacitation components lead to substantial behavioral changes and that farmers’ responsiveness to sanctions coincides with the overall commitment to forest law enforcement. There is no evidence of spatial displacement or attempts to elude monitoring. In a counterfactual scenario without sanctions, farmers’ deforestation would increase by 29%. These findings suggest that general deterrence can make environmental sanctions a powerful tool for combating deforestation and promoting reforestation at scale.

This paper proposes a novel instrumental variable to estimate the causal impact of law enforcement on deforestation. DETER, a satellite-based system for mapping land cover, is the key tool in Amazon monitoring. It determines the location of recent forest clearings and is used to target enforcement. Because DETER cannot detect land cover patterns beneath clouds, it detects no clearings in covered areas. Results conrm that DETER cloud coverage is systematically correlated with nes, a proxy for the presence of law enforcers. The study explores this exogenous source of variation in the allocation of law enforcers as an instrument for the intensity of enforcement. Stricter law enforcement eectively deterred Amazon deforestation, helping avoid over 22,000 km2 of cleared forest area per sample year. Leakage of criminal activity into neighboring areas does not appear to have occurred. Results are submitted to a series of robustness checks.